Using a pre-trained model to detect objects in an image.
import numpy as np
import os
import sys
import tensorflow as tf
import time
from collections import defaultdict
from io import StringIO
from matplotlib import pyplot as plt
from PIL import Image
%matplotlib inline
Here are the imports from the object detection module.
from object_detection.utils import label_map_util
from object_detection.utils import visualization_utils as vis_util
ssd_inception_sim_model = 'frozen_models/frozen_sim_inception/frozen_inference_graph.pb'
ssd_inception_real_model = 'frozen_models/frozen_real_inception_6561/frozen_inference_graph.pb'
faster_rcnn_sim_model = 'frozen_models/faster_rcnn_frozen_sim/frozen_inference_graph.pb'
faster_rcnn_real_model = 'frozen_models/faster_rcnn_frozen_real/frozen_inference_graph.pb'
PATH_TO_LABELS = 'label_map.pbtxt'
NUM_CLASSES = 14
Label maps map indices to category names, so that when our convolution network predicts 2, we know that this corresponds to Red. Here we use internal utility functions, but anything that returns a dictionary mapping integers to appropriate string labels would be fine.
label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
category_index = label_map_util.create_category_index(categories)
print(category_index)
def load_image_into_numpy_array(image):
(im_width, im_height) = image.size
return np.array(image.getdata()).reshape(
(im_height, im_width, 3)).astype(np.uint8)
from glob import glob
detection_graph = tf.Graph()
with detection_graph.as_default():
od_graph_def = tf.GraphDef()
with tf.gfile.GFile(ssd_inception_sim_model, 'rb') as fid:
serialized_graph = fid.read()
od_graph_def.ParseFromString(serialized_graph)
tf.import_graph_def(od_graph_def, name='')
PATH_TO_TEST_IMAGES_DIR = 'test_images_sim'
print(os.path.join(PATH_TO_TEST_IMAGES_DIR, '*.jpg'))
TEST_IMAGE_PATHS = glob(os.path.join(PATH_TO_TEST_IMAGES_DIR, '*.jpg'))
print("Length of test images:", len(TEST_IMAGE_PATHS))
# Size, in inches, of the output images.
IMAGE_SIZE = (12, 8)
with detection_graph.as_default():
with tf.Session(graph=detection_graph) as sess:
# Definite input and output Tensors for detection_graph
image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
# Each box represents a part of the image where a particular object was detected.
detection_boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
# Each score represent how level of confidence for each of the objects.
# Score is shown on the result image, together with the class label.
detection_scores = detection_graph.get_tensor_by_name('detection_scores:0')
detection_classes = detection_graph.get_tensor_by_name('detection_classes:0')
num_detections = detection_graph.get_tensor_by_name('num_detections:0')
for image_path in TEST_IMAGE_PATHS:
image = Image.open(image_path)
# the array based representation of the image will be used later in order to prepare the
# result image with boxes and labels on it.
image_np = load_image_into_numpy_array(image)
# Expand dimensions since the model expects images to have shape: [1, None, None, 3]
image_np_expanded = np.expand_dims(image_np, axis=0)
time0 = time.time()
# Actual detection.
(boxes, scores, classes, num) = sess.run(
[detection_boxes, detection_scores, detection_classes, num_detections],
feed_dict={image_tensor: image_np_expanded})
time1 = time.time()
boxes = np.squeeze(boxes)
scores = np.squeeze(scores)
classes = np.squeeze(classes).astype(np.int32)
# Visualization of the results of a detection.
vis_util.visualize_boxes_and_labels_on_image_array(
image_np, boxes, classes, scores,
category_index,
use_normalized_coordinates=True,
line_thickness=6)
plt.figure(figsize=IMAGE_SIZE)
plt.imshow(image_np)
plt.show()
min_score_thresh = .50
for i in range(boxes.shape[0]):
if scores is None or scores[i] > min_score_thresh:
class_name = category_index[classes[i]]['name']
print('{}'.format(class_name), scores[i])
fx = 0.97428
fy = 1.73205
perceived_width_x = (boxes[i][3] - boxes[i][1]) * 800
perceived_width_y = (boxes[i][2] - boxes[i][0]) * 600
# ymin, xmin, ymax, xmax = box
# depth_prime = (width_real * focal) / perceived_width
perceived_depth_x = ((.1 * fx) / perceived_width_x)
perceived_depth_y = ((.3 * fy) / perceived_width_y )
estimated_distance = round((perceived_depth_x + perceived_depth_y) / 2)
print("Distance (metres)", estimated_distance)
print("Time in milliseconds", (time1 - time0) * 1000, "\n")
detection_graph = tf.Graph()
with detection_graph.as_default():
od_graph_def = tf.GraphDef()
with tf.gfile.GFile(ssd_inception_real_model, 'rb') as fid:
serialized_graph = fid.read()
od_graph_def.ParseFromString(serialized_graph)
tf.import_graph_def(od_graph_def, name='')
PATH_TO_TEST_IMAGES_DIR = 'test_images_udacity'
print(os.path.join(PATH_TO_TEST_IMAGES_DIR, '*.jpg'))
TEST_IMAGE_PATHS = glob(os.path.join(PATH_TO_TEST_IMAGES_DIR, '*.jpg'))
print("Length of test images:", len(TEST_IMAGE_PATHS))
# Size, in inches, of the output images.
IMAGE_SIZE = (12, 8)
with detection_graph.as_default():
with tf.Session(graph=detection_graph) as sess:
# Definite input and output Tensors for detection_graph
image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
# Each box represents a part of the image where a particular object was detected.
detection_boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
# Each score represent how level of confidence for each of the objects.
# Score is shown on the result image, together with the class label.
detection_scores = detection_graph.get_tensor_by_name('detection_scores:0')
detection_classes = detection_graph.get_tensor_by_name('detection_classes:0')
num_detections = detection_graph.get_tensor_by_name('num_detections:0')
for image_path in TEST_IMAGE_PATHS:
image = Image.open(image_path)
# the array based representation of the image will be used later in order to prepare the
# result image with boxes and labels on it.
image_np = load_image_into_numpy_array(image)
# Expand dimensions since the model expects images to have shape: [1, None, None, 3]
image_np_expanded = np.expand_dims(image_np, axis=0)
time0 = time.time()
# Actual detection.
(boxes, scores, classes, num) = sess.run(
[detection_boxes, detection_scores, detection_classes, num_detections],
feed_dict={image_tensor: image_np_expanded})
time1 = time.time()
boxes = np.squeeze(boxes)
scores = np.squeeze(scores)
classes = np.squeeze(classes).astype(np.int32)
# Visualization of the results of a detection.
vis_util.visualize_boxes_and_labels_on_image_array(
image_np, boxes, classes, scores,
category_index,
use_normalized_coordinates=True,
line_thickness=6)
plt.figure(figsize=IMAGE_SIZE)
plt.imshow(image_np)
plt.show()
min_score_thresh = .50
for i in range(boxes.shape[0]):
if scores is None or scores[i] > min_score_thresh:
class_name = category_index[classes[i]]['name']
print('{}'.format(class_name), scores[i])
fx = 1345.200806
fy = 1353.838257
perceived_width_x = (boxes[i][3] - boxes[i][1]) * 800
perceived_width_y = (boxes[i][2] - boxes[i][0]) * 600
# ymin, xmin, ymax, xmax = box
# depth_prime = (width_real * focal) / perceived_width
perceived_depth_x = ((.1 * fx) / perceived_width_x)
perceived_depth_y = ((.3 * fy) / perceived_width_y )
estimated_distance = round((perceived_depth_x + perceived_depth_y) / 2)
print("Distance (metres)", estimated_distance)
print("Time in milliseconds", (time1 - time0) * 1000, "\n")
detection_graph = tf.Graph()
with detection_graph.as_default():
od_graph_def = tf.GraphDef()
with tf.gfile.GFile(faster_rcnn_sim_model, 'rb') as fid:
serialized_graph = fid.read()
od_graph_def.ParseFromString(serialized_graph)
tf.import_graph_def(od_graph_def, name='')
PATH_TO_TEST_IMAGES_DIR = 'test_images_sim'
print(os.path.join(PATH_TO_TEST_IMAGES_DIR, '*.jpg'))
TEST_IMAGE_PATHS = glob(os.path.join(PATH_TO_TEST_IMAGES_DIR, '*.jpg'))
print("Length of test images:", len(TEST_IMAGE_PATHS))
# Size, in inches, of the output images.
IMAGE_SIZE = (12, 8)
with detection_graph.as_default():
with tf.Session(graph=detection_graph) as sess:
# Definite input and output Tensors for detection_graph
image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
# Each box represents a part of the image where a particular object was detected.
detection_boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
# Each score represent how level of confidence for each of the objects.
# Score is shown on the result image, together with the class label.
detection_scores = detection_graph.get_tensor_by_name('detection_scores:0')
detection_classes = detection_graph.get_tensor_by_name('detection_classes:0')
num_detections = detection_graph.get_tensor_by_name('num_detections:0')
for image_path in TEST_IMAGE_PATHS:
image = Image.open(image_path)
# the array based representation of the image will be used later in order to prepare the
# result image with boxes and labels on it.
image_np = load_image_into_numpy_array(image)
# Expand dimensions since the model expects images to have shape: [1, None, None, 3]
image_np_expanded = np.expand_dims(image_np, axis=0)
time0 = time.time()
# Actual detection.
(boxes, scores, classes, num) = sess.run(
[detection_boxes, detection_scores, detection_classes, num_detections],
feed_dict={image_tensor: image_np_expanded})
time1 = time.time()
boxes = np.squeeze(boxes)
scores = np.squeeze(scores)
classes = np.squeeze(classes).astype(np.int32)
# Visualization of the results of a detection.
vis_util.visualize_boxes_and_labels_on_image_array(
image_np, boxes, classes, scores,
category_index,
use_normalized_coordinates=True,
line_thickness=6)
plt.figure(figsize=IMAGE_SIZE)
plt.imshow(image_np)
plt.show()
min_score_thresh = .50
for i in range(boxes.shape[0]):
if scores is None or scores[i] > min_score_thresh:
class_name = category_index[classes[i]]['name']
print('{}'.format(class_name), scores[i])
fx = 0.97428
fy = 1.73205
perceived_width_x = (boxes[i][3] - boxes[i][1]) * 800
perceived_width_y = (boxes[i][2] - boxes[i][0]) * 600
# ymin, xmin, ymax, xmax = box
# depth_prime = (width_real * focal) / perceived_width
perceived_depth_x = ((.1 * fx) / perceived_width_x)
perceived_depth_y = ((.3 * fy) / perceived_width_y )
estimated_distance = round((perceived_depth_x + perceived_depth_y) / 2)
print("Distance (metres)", estimated_distance)
print("Time in milliseconds", (time1 - time0) * 1000, "\n")
detection_graph = tf.Graph()
with detection_graph.as_default():
od_graph_def = tf.GraphDef()
with tf.gfile.GFile(faster_rcnn_real_model, 'rb') as fid:
serialized_graph = fid.read()
od_graph_def.ParseFromString(serialized_graph)
tf.import_graph_def(od_graph_def, name='')
PATH_TO_TEST_IMAGES_DIR = 'test_images_udacity'
print(os.path.join(PATH_TO_TEST_IMAGES_DIR, '*.jpg'))
TEST_IMAGE_PATHS = glob(os.path.join(PATH_TO_TEST_IMAGES_DIR, '*.jpg'))
print("Length of test images:", len(TEST_IMAGE_PATHS))
# Size, in inches, of the output images.
IMAGE_SIZE = (12, 8)
with detection_graph.as_default():
with tf.Session(graph=detection_graph) as sess:
# Definite input and output Tensors for detection_graph
image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
# Each box represents a part of the image where a particular object was detected.
detection_boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
# Each score represent how level of confidence for each of the objects.
# Score is shown on the result image, together with the class label.
detection_scores = detection_graph.get_tensor_by_name('detection_scores:0')
detection_classes = detection_graph.get_tensor_by_name('detection_classes:0')
num_detections = detection_graph.get_tensor_by_name('num_detections:0')
for image_path in TEST_IMAGE_PATHS:
image = Image.open(image_path)
# the array based representation of the image will be used later in order to prepare the
# result image with boxes and labels on it.
image_np = load_image_into_numpy_array(image)
# Expand dimensions since the model expects images to have shape: [1, None, None, 3]
image_np_expanded = np.expand_dims(image_np, axis=0)
time0 = time.time()
# Actual detection.
(boxes, scores, classes, num) = sess.run(
[detection_boxes, detection_scores, detection_classes, num_detections],
feed_dict={image_tensor: image_np_expanded})
time1 = time.time()
boxes = np.squeeze(boxes)
scores = np.squeeze(scores)
classes = np.squeeze(classes).astype(np.int32)
# Visualization of the results of a detection.
vis_util.visualize_boxes_and_labels_on_image_array(
image_np, boxes, classes, scores,
category_index,
use_normalized_coordinates=True,
line_thickness=6)
plt.figure(figsize=IMAGE_SIZE)
plt.imshow(image_np)
plt.show()
min_score_thresh = .50
for i in range(boxes.shape[0]):
if scores is None or scores[i] > min_score_thresh:
class_name = category_index[classes[i]]['name']
print('{}'.format(class_name), scores[i])
fx = 1345.200806
fy = 1353.838257
perceived_width_x = (boxes[i][3] - boxes[i][1]) * 800
perceived_width_y = (boxes[i][2] - boxes[i][0]) * 600
# ymin, xmin, ymax, xmax = box
# depth_prime = (width_real * focal) / perceived_width
perceived_depth_x = ((.1 * fx) / perceived_width_x)
perceived_depth_y = ((.3 * fy) / perceived_width_y )
estimated_distance = round((perceived_depth_x + perceived_depth_y) / 2)
print("Distance (metres)", estimated_distance)
print("Time in milliseconds", (time1 - time0) * 1000, "\n")